A preventive maintenance method for distributed energy supply system equipment based on entropy theory

Abstract

A preventive maintenance method for distributed energy supply system equipment based on entropy theory is proposed. An equipment evaluation model based on Latin sampling method is established to evaluate the importance and hazard of each subsystem and related components of distributed energy supply equipment. Ten operation and maintenance importance influencing factors and four operation and maintenance hazard evaluation factors are created. The influence degree of each factor is quantitatively described through the hierarchical analysis method. Combined with the dynamic statistical analysis of fault data, the influence weight of each factor is corrected in a closed loop to obtain the scoring standard and weight of the corresponding influencing factors of each subsystem and component of distributed energy supply equipment. Combined with quantitative indicators such as equipment maintenance cost, equipment importance, and fault hazard, a preventive maintenance method for distributed energy supply system equipment is proposed based on entropy theory to eliminate the dimensional influence between different indicators, mine multi-layer evaluation indicator information, determine scientific and reasonable maintenance time and cycle, and solve the problem of over-maintenance and under-maintenance of distributed energy supply equipment.

Description

Technical Field

[0001] The invention relates to a preventive maintenance method for equipment of a distributed energy supply system. Background Art

[0002] As the national clean energy development strategy continues to advance, the distributed energy supply system industry continues to develop, the types of distributed energy supply equipment increase, and the complexity continues to increase. How to perform equipment maintenance work to make the equipment reliable and economical has become a research hotspot. Distributed energy supply equipment will gradually degrade during long-term operation. The system components are complex and expensive, and the various parts of the system are prone to chain effects. The consequences of damage or failure may be very serious.

[0003] However, the current maintenance of distributed energy supply equipment is overly dependent on the experience of operation and maintenance staff and the data provided by the original manufacturer, and has poor decision-making for specific problems occurring on site. On the other hand, the traditional reliability-centered maintenance method does not consider the impact of different influencing factors on the maintenance strategy, and ignores the changes in maintenance decision-making corresponding to changes in equipment operation time. Therefore, the traditional reliability-centered maintenance method cannot fully meet the actual maintenance work needs.

[0004] Preventive maintenance helps keep equipment running normally through regular and routine maintenance, thus avoiding the costly losses caused by unplanned downtime or failure. Therefore, in order to ensure the normal operation of the system and avoid the losses caused by production stoppage accidents, preventive maintenance is extremely important. Summary of the invention

[0005] The purpose of the present invention is to overcome the shortcomings of the prior art and propose a reliability-centered preventive maintenance method for distributed energy supply systems based on entropy theory. The present invention applies entropy theory to optimize the reliability-centered maintenance decision model of distributed energy supply systems, and formulates more scientific and reasonable maintenance timing and cycle according to engineering practice, thereby solving the over-maintenance and under-maintenance phenomenon of distributed energy supply equipment caused by unreasonable maintenance timing and cycle.

[0006] The invention discloses a method for the preventive maintenance of distributed energy supply system equipment based on entropy theory. According to the actual operation and maintenance experience, the power market demand and the actual situation and characteristics of the equipment, an equipment evaluation model based on the Latin sampling method is established to evaluate the importance and criticality of each subsystem and related components of the distributed energy supply equipment. Ten operation and maintenance importance influencing factors and four operation and maintenance criticality evaluation factors are created. The influence degree of each factor is quantitatively described by the hierarchical analysis method. Combined with the dynamic statistical analysis of the fault data, the influence weight of each factor is corrected in a closed loop to obtain the scoring standard and weight of the corresponding influencing factors of each subsystem and component of the distributed energy supply equipment. On this basis, the equipment maintenance cost is introduced, and a preventive maintenance cost model based on the Weibull distribution is established. Combined with quantitative indicators such as equipment maintenance cost, equipment importance, and fault criticality, a preventive maintenance method for distributed energy supply system equipment based on entropy theory is proposed to determine a more scientific and reasonable maintenance time and cycle. The specific steps are as follows:

[0007] Step 1: Establish a device importance model based on Latin sampling method

[0008] Based on actual operation and maintenance experience, in accordance with the power market demand and the actual situation and characteristics of the equipment, ten influencing factors of equipment operation and maintenance are established: the impact of the failure of distributed energy supply equipment on the safety of operation and maintenance personnel and the environment, the impact of the failure of distributed energy supply equipment on the system function, the impact of equipment failure on maintenance costs, the impact of failure on downtime losses, the impact of equipment status monitorability, the impact of downtime, the impact of the difficulty of inspection and maintenance, the impact of failure frequency, the impact of the timeliness of spare parts supply, and the impact of external environment changes on the degree of failure of distributed energy supply equipment.

[0009] In the analysis of various influencing factors, in order to reduce the complexity of equipment importance analysis, the hierarchical analysis method is used. At the same time, the above ten influencing factors are divided into 5 levels and given interval scores, so as to quantify the scoring criteria and weights of the corresponding influencing factors of each subsystem and component of the distributed energy supply equipment, specifically:

[0010] First, construct a judgment matrix. Analyze the importance of the ten operation and maintenance evaluation factors. The relative importance of each factor can be expressed by the following judgment matrix U:

[0011]

[0012] The element u in the matrix represents the relative importance of the evaluation factor.

[0013] Then calculate the importance of the equipment and sort it. Calculate the characteristic root of matrix A, λ is the corresponding eigenvalue of the matrix, and take the largest eigenvalue λ max Substitute into the homogeneous linear equations (2):

[0014]

[0015] The weights of each influencing factor are solved as w 1 ,w 2 ,w 3 ,w 4 ,w 5 ,w 6 ,w 7 ,w 8 ,w 9 ,w 0 , get the maximum eigenvalue λ max The corresponding eigenvector W;

[0016] W=(w 1 ,w 2 ,...,w 0 ) T (3)

[0017] In formula (3), T is the transpose symbol of the matrix; the weights of the ten influencing factors are obtained, thereby quantifying the impact of each factor. The priority of each influencing factor is calculated and sorted according to the weight.

[0018] Finally, a consistency check is performed, as shown in formula (4):

[0019] CR=CI / RI (4)

[0020] Where CR is the random consistency ratio of the judgment matrix; CI is the general consistency index of the judgment matrix, and its value is: CI = (λ max -n) / (n-1); RI is the average random consistency index of the judgment matrix. The RI values ​​are shown in the following table, and n is the number of importance evaluation factors.

[0021] n 1 2 3 4 5 n 6 7 8 9 10 RI 0 0 0.52 0.89 1.12 RI 1.26 1.36 1.41 1.46 1.49

[0022] When CR is less than 0.1, it is considered that the calculation results of the current judgment matrix have satisfactory consistency, indicating that the weight distribution of different factors is reasonable; otherwise, it is necessary to adjust the judgment matrix, such as applying the judgment of the inverse M-matrix and the parallel algorithm, and recalculate.

[0023] When performing calculations, the application of Latin sampling can dynamically and closed-loop correct the weights of each importance evaluation factor, so that the impact of the evaluation factors is more in line with the actual working conditions of the equipment, enhance the robustness of the importance of each subsystem and component of the equipment, and reduce the impact of human interference on the importance evaluation.

[0024] Latin sampling is used to divide the input probability distribution into several independent segments, each of which has the same probability of being selected, and samples are randomly drawn from each segment. Latin sampling is used to statistically analyze the importance weights of each subsystem and component of the equipment, and the importance weights of each subsystem and component are effectively ranked.

[0025] On the basis of determining the importance weight of the equipment, a linear weighted mathematical model is used to calculate the importance evaluation index of each subsystem and component of the equipment. The calculation formula is:

[0026]

[0027] In the formula, Index is the importance evaluation index, a i is the weight of the i-th factor; m i is the score of the i-th factor, and the subscript i is the serial number of the influencing factor.

[0028] Step 2: Establish a device criticality model based on the Latin sampling method

[0029] The hazard degree of distributed energy supply equipment failure is calculated by using the hierarchical analysis method combined with Latin sampling. Four hazard influencing factors of equipment operation and maintenance are determined: the degree of distributed energy supply equipment failure, the degree of impact of equipment historical failures, the probability of equipment failure, and the severity of equipment failure impact.

[0030] The above four influencing factors are divided into five levels and given interval scores, so as to quantify the scoring criteria and weights of the corresponding influencing factors of each subsystem and component of the distributed energy supply equipment.

[0031] First, construct a judgment matrix. The relative severity of the four severity influencing factors is expressed by the following judgment matrix B:

[0032]

[0033] The element b in the matrix represents the relative severity of the evaluation factor.

[0034] Then calculate the equipment hazard level and sort it. Calculate the maximum characteristic root λ of the judgment matrix B max , substitute into the homogeneous linear equations (7):

[0035]

[0036] The weights of each influencing factor are solved as x 1 ,x 2 ,x 3 ,x 4 , get the maximum characteristic root λ max The corresponding eigenvector X is:

[0037] X=(x 1 ,x 2 ,x 3 ,x 4 ) T (8)

[0038] The weight of each influencing factor can be obtained. Thus, the impact of each factor can be quantified. The priority of each influencing factor can be calculated and sorted according to the weight.

[0039] Finally, a consistency check is performed, as shown in formula (9):

[0040] CR=CI / RI (9)

[0041] In the formula, CR is called the random consistency ratio of the judgment matrix; CI is called the general consistency index of the judgment matrix, and its value is CI = (λ max -n) / (n-1); RI is called the average random consistency index of the judgment matrix, and its value is as shown in the following table.

[0042] n 1 2 3 4 RI 0 0 0.52 0.89

[0043] When CR is less than 0.1, it is considered that the calculation results of the current judgment matrix have satisfactory consistency, indicating that the weight distribution of different factors is reasonable; otherwise, the judgment matrix needs to be adjusted and recalculated.

[0044] When performing calculations, the application of Latin sampling can dynamically and closed-loop correct the weights of each hazard assessment factor, so that the impact of the assessment factors is more in line with the actual working conditions of the equipment, enhances the robustness of the importance of each subsystem and component of the equipment, and reduces the impact of human interference on the hazard assessment.

[0045] Latin sampling is used to divide the input probability distribution into several independent segments, each of which has the same probability of being selected, and samples are randomly drawn from each segment. Latin sampling is used to statistically analyze the criticality weights of each subsystem and component of the equipment, and the criticality weights of each subsystem and component are effectively ranked.

[0046] On the basis of determining the criticality weight of the equipment, a linear weighted mathematical model is used to calculate the criticality evaluation index of each subsystem and component of the equipment. The calculation formula is:

[0047]

[0048] In the formula, Index2 is the hazard evaluation index, f j is the weight of the jth hazard assessment factor; g j is the score of the jth factor, and the subscript j is the serial number of the influencing factor.

[0049] Step 3: Establish a preventive maintenance method for distributed energy supply system equipment based on entropy theory

[0050] Taking into account quantitative indicators such as component maintenance cost, importance and criticality, a preventive maintenance method for distributed energy supply system equipment is established based on entropy theory, and the optimal maintenance period S is obtained, which ensures the economy, reliability and availability of distributed energy supply system equipment.

[0051] The preventive maintenance model of distributed energy supply system equipment based on entropy theory is shown in formula (11):

[0052] I n (S) = e 1 (S)z n1 (S)+e 2 (S)z n2 +e 3 (S)z n3 (11)

[0053] In the formula, Z n1 (S) is the maintenance cost of the nth component corresponding to the cost-optimal model at time S; Z n2 is the corresponding importance of the nth component determined by the Latin sampling method, Z n3 is the equipment hazard level; e 1 (S), e 2 (S), e 3 (S) are the weights of maintenance cost, importance and criticality in the cost-optimal model.

[0054] The cost optimization model is as follows:

[0055]

[0056] In the formula, c 1 The loss of subsequent repairs; 2 is the periodic maintenance loss; F(s) is the cumulative failure probability function; R(s) is the reliability function. It is easy to know that Z n1 When (S) is minimum, the cost of preventive maintenance is the lowest, and the corresponding time S is the economically optimal maintenance cycle.

[0057] According to the entropy theory, in formula (11),

[0058] k=1 / ln(n) (13)

[0059]

[0060]

[0061]

[0062] Where k is the calculation coefficient; P n1(S) is the proportion of the maintenance cost of the nth component at time S in the sum of importance, criticality and maintenance cost; P n2 (S) is the proportion of the importance of the nth component at time S in the sum of importance, criticality and maintenance cost, P n3 (S) is the proportion of the criticality of the nth component at time S in the sum of importance, criticality and maintenance cost.

[0063]

[0064]

[0065]

[0066] e 1 (S) = (1-Q 1 (S)) / [(1-Q 1 (S))+(1-Q 2 (S))+(1-Q 3 (S))] (20)

[0067] e 2 (S) = (1-Q 2 (S)) / [(1-Q 1 (S))+(1-Q 2 (S))+(1-Q 3 (S))] (21)

[0068] e 3 (S) = (1-Q 3 (S)) / [(1-Q 1 (S))+(1-Q 2 (S))+(1-Q 3 (S))] (22)

[0069] In the formula, Q 1 (S), Q 2 (S), Q 3 (S) corresponds to e 1 (S), e 2 (S), e 3 (S) Entropy of each risk factor.

[0070] In this way, the time S corresponding to the minimum value of the equipment maintenance decision index I(S) is obtained, so as to obtain the optimal maintenance time interval S that ensures economy, reliability and availability. BRIEF DESCRIPTION OF THE DRAWINGS

[0071] Picture 1 A flow chart of a method for preventive maintenance of distributed energy supply equipment based on entropy theory of the present invention. DETAILED DESCRIPTION

[0072] The present invention is further described in detail below with reference to the accompanying drawings and embodiments.

[0073] like Picture 1 As shown, an embodiment of the method for preventive maintenance of distributed energy supply system equipment based on entropy theory of the present invention includes the following steps:

[0074] 1. Establish a device importance model based on Latin sampling method

[0075] According to the actual operation and maintenance experience, in accordance with the power market demand and the actual situation and characteristics of the equipment, ten important factors affecting equipment operation and maintenance are created: the impact of the failure of distributed energy supply equipment on the safety of operation and maintenance personnel and the environment, the impact of the failure of distributed energy supply equipment on the system function, the impact of equipment failure on maintenance costs, the impact of failure on downtime losses, the impact of equipment status monitorability, the impact of downtime, the impact of maintenance difficulty, the impact of failure frequency, the impact of spare parts supply timeliness, and the impact of external environment changes on the degree of failure of distributed energy supply equipment. In the analysis of various influencing factors, in order to reduce the complexity of equipment importance analysis, hierarchical analysis is used for analysis. The above ten influencing factors are divided into 5 levels and given interval scores, so as to quantify the scoring criteria and weights of the corresponding influencing factors of each subsystem and component of the distributed energy supply equipment.

[0076]

[0077] In the formula, A is the importance matrix, the element u in the matrix represents the relative importance of the evaluation factor; λ is the corresponding eigenvalue of the matrix; w is the importance weight; CR is the consistency ratio of the judgment matrix; CI is the general consistency index of the judgment matrix; RI is the average random consistency index of the judgment matrix; Index is the importance evaluation index, a i is the weight of the i-th factor; m i is the score of the i-th factor, and the subscript i is the serial number of the influencing factor.

[0078] 2. Establish equipment criticality model based on Latin sampling method

[0079] The hazard level of distributed energy supply equipment failure is calculated by using the hierarchical analysis method combined with Latin sampling. Four hazard influencing factors of equipment operation and maintenance are determined: the degree of distributed energy supply equipment failure, the degree of impact of equipment historical failures, the probability of equipment failure, and the severity of equipment failure impact. The above four influencing factors are divided into 5 levels and given interval scores, so as to quantify the scoring criteria and weights of the corresponding influencing factors of each subsystem and component of the distributed energy supply equipment.

[0080]

[0081] In the formula, B is the hazard matrix; the element b in the matrix represents the relative hazard of the evaluation factor; x is the hazard weight; CR is the consistency ratio of the judgment matrix; CI is the general consistency index of the judgment matrix; RI is the average random consistency index of the judgment matrix; Index2 is the hazard evaluation index, f i is the weight of the i-th factor; g i is the score of the i-th factor, and the subscript j is the serial number of the influencing factor.

[0082] 3. Establish a preventive maintenance method for distributed energy supply system equipment based on entropy theory

[0083] Based on the entropy theory, a preventive maintenance method for distributed energy supply system equipment is proposed. Considering quantitative indicators such as component maintenance cost, importance and fault hazard, the optimal maintenance cycle S is obtained, which ensures the economy, reliability and availability of distributed energy supply system equipment.

[0084]

[0085] In the formula, Z n1 (S) is the maintenance cost of the nth component corresponding to the cost-optimal model at time S; Z n2 is the corresponding importance of the nth component determined by the Latin sampling method; Z n3 is the equipment hazard level; e 1 (S), e 2 (S), e 3 (S) are the weights of the corresponding factors; c 1 The loss of subsequent repairs; 2 is the periodic maintenance loss; F(s) is the cumulative failure probability function; R(s) is the reliability function; k is the calculation coefficient; P n1 (S) is the proportion of the maintenance cost of the nth component at time S in the sum of importance, criticality and maintenance cost; P n2 (S) is the proportion of the importance of the nth component at time S in the sum of importance, criticality and maintenance cost; P n3 (S) is the proportion of the criticality of the nth component at time S in the sum of importance, criticality and maintenance cost; Q 1 (S), Q 2 (S), Q 3 (S) corresponds to e 1 (S), e 2 (S), e 3 (S) Entropy of each risk factor.

Claims

1. A method for preventive maintenance of distributed energy supply system equipment based on entropy theory, characterized in that: The method is based on actual operation and maintenance experience, according to the power market demand and the actual situation and characteristics of the equipment, and establishes an equipment evaluation based on the Latin sampling method. The model evaluates the importance and hazard of each subsystem and related components of the distributed energy supply equipment; creates ten operation and maintenance importance influencing factors and four operation and maintenance hazard evaluation factors, and quantitatively describes the degree of influence of each factor through the hierarchical analysis method. Combined with the dynamic statistical analysis of fault data, the influence weight of each factor is corrected in a closed loop to obtain the scoring standard and weight of the corresponding influencing factors of each subsystem and component of the distributed energy supply equipment; combined with the equipment maintenance cost, equipment importance, and fault hazard quantification indicators, a preventive maintenance method for distributed energy supply system equipment is proposed based on the entropy theory to determine the maintenance opportunity and cycle; the steps are as follows: Step 1: Establish an equipment importance model based on the Latin sampling method; the ten importance influencing factors of equipment operation and maintenance are: the impact of the failure of distributed energy supply equipment on the safety of operation and maintenance personnel and the environment, the impact of the failure of distributed energy supply equipment on the system function, the impact of equipment failure on maintenance costs, the impact of failure on downtime losses, the impact of equipment status monitorability, the impact of downtime, the impact of maintenance difficulty, the impact of failure frequency, the impact of spare parts supply timeliness, and the impact of external environment changes on the degree of failure of distributed energy supply equipment; The ten influencing factors are divided into five levels and given interval scores, so as to quantify the scoring criteria and weights of the corresponding influencing factors of each subsystem and component of the distributed energy supply equipment; Step 2: Establish an equipment hazard model based on the Latin sampling method; use the analytic hierarchy process combined with Latin sampling to calculate the hazard level of distributed energy supply equipment failures; determine the four hazard influencing factors of equipment operation and maintenance: the degree of distributed energy supply equipment failure, the degree of impact of equipment historical failures, the probability of equipment failure, and the severity of equipment failure impact; The above four influencing factors are divided into 5 levels and given interval scores, so as to quantify the scoring criteria and weights of the corresponding influencing factors of each subsystem and component of the distributed energy supply equipment; Step 3: Establish a preventive maintenance method for distributed energy supply system equipment based on entropy theory; Taking into account the maintenance cost, importance and hazard quantification indicators of components, a preventive maintenance method for distributed energy supply system equipment is established based on entropy theory to obtain the optimal maintenance cycle S. The preventive maintenance model of distributed energy supply system equipment based on entropy theory is shown in formula (11): （11） In the formula, Z n1 (S) is the maintenance cost of the nth component corresponding to the cost-optimal model at time S; Z n2 is the corresponding importance of the nth component determined by the Latin sampling method, Z n3 is the corresponding hazard level of the nth component determined by the Latin sampling method; e1(S), e2(S), and e3(S) are the weights of the corresponding factors respectively; The cost optimization model is as follows: （12） In the formula, c1 is the post-maintenance loss, c2 is the periodic maintenance loss, F(s) is the cumulative failure probability function, and R(s) is the reliability function. It is easy to know that Z n1 When (S) is minimum, the cost of preventive maintenance is the lowest, and the corresponding time S is the economically optimal maintenance cycle; According to the entropy theory, in formula (11), （13） （14） （15） （16） Where k is the calculation coefficient; P n1 (S) is the proportion of the maintenance cost of the nth component at time S in the sum of importance, criticality and maintenance cost; P n2 (S) is the proportion of the importance of the nth component at time S in the sum of importance, criticality and maintenance cost, P n3 (S) is the proportion of the criticality of the nth component at time S in the sum of importance, criticality and maintenance cost; （17） （18） （19） （20） （21） （22） Where Q1(S), Q2(S), and Q3(S) are the entropies of the risk factors corresponding to e1(S), e2(S), and e3(S), respectively; In this way, the time S corresponding to the minimum value of the equipment maintenance decision index I (S) is obtained, and the optimal maintenance time interval S is obtained.

2. According to claim 1, a method for preventive maintenance of distributed energy supply system equipment based on entropy theory is characterized in that: The step 1 includes: first constructing a judgment matrix; the relative importance of the ten operation and maintenance importance influencing factors is represented by the following judgment matrix U: （1） The element u in the matrix represents the relative importance of the evaluation factor; Then calculate the importance of the equipment and sort it; calculate the maximum characteristic root λ of the judgment matrix U max , substitute into the homogeneous linear equations (2): （2） The weights of each influencing factor are solved as w1, w2, w3, w4, w5, w6, w7, w8, w9, w0, and λ is the matrix eigenvalue. The maximum eigenvalue λ max The corresponding eigenvector W is shown in formula (3): （3） In formula (3), T is the transpose symbol of the matrix; the weights of the ten influencing factors are obtained; the priority ranking of each influencing factor is calculated according to the weight size; Finally, a consistency check is performed, as shown in formula (4): （4） Where CR is the random consistency ratio of the judgment matrix; CI is the general consistency index of the judgment matrix, and its value is CI=(λ max -n) / (n-1); RI is the average random consistency index of the judgment matrix; When CR is less than 0.1, it is considered that the current judgment matrix calculation results have satisfactory consistency, indicating that the weight distribution of different factors is reasonable; otherwise, the judgment matrix needs to be adjusted and recalculated; During the calculation, the Latin sampling dynamic closed loop is applied to correct the weight of each importance evaluation factor, so that the impact of the evaluation factor is more in line with the actual working conditions of the equipment, the robustness of the importance of each subsystem and component of the equipment is enhanced, and the impact of human interference on the importance evaluation is reduced; Use Latin sampling to divide the input probability distribution into several independent segments, each segment has the same probability of being selected, and randomly select samples from each segment; calculate the importance weights of each subsystem and component of the equipment through Latin sampling, and sort the importance weights of each subsystem and component; On the basis of determining the importance weight of the equipment, a linear weighted mathematical model is used to calculate the importance evaluation index of each subsystem and component of the equipment. The calculation formula is: （5） In the formula, Index is the importance evaluation index, a i is the weight of the i-th factor; m i is the score of the i-th factor, and the subscript i is the serial number of the influencing factor.

3. The method for preventive maintenance of distributed energy supply system equipment based on entropy theory according to claim 1 is characterized in that: The step 2 includes: first constructing a judgment matrix. The criticality analysis has four operation and maintenance evaluation factors. The relative criticality between each factor can be represented by the following judgment matrix B: （6） The element b in the matrix represents the relative severity of the evaluation factor; Then calculate the equipment hazard level and sort it; calculate the maximum characteristic root λ of the judgment matrix B max , substitute into the homogeneous linear equations (7): （7） The weights of each influencing factor are solved as x1, x2, x3, x4, then the maximum characteristic root λ max The corresponding eigenvector X is: （8） Obtain the weight matrix X corresponding to each influencing factor; calculate and sort the priority of each influencing factor according to the weight; Finally, a consistency check is performed, as shown in formula (9): （9） Where CR is the random consistency ratio of the judgment matrix; CI is the general consistency index of the judgment matrix, and its value is CI=(λ max -n) / (n-1); RI is the average random consistency index of the judgment matrix; When CR is less than 0.1, it is considered that the current judgment matrix calculation results have satisfactory consistency, indicating that the weight distribution of different factors is reasonable; otherwise, the judgment matrix needs to be adjusted and recalculated; When performing calculations, Latin sampling dynamic closed loop is used to correct the weights of each hazard evaluation factor, so that the impact of the evaluation factor is more in line with the actual working conditions of the equipment; Use Latin sampling to divide the input probability distribution into several independent segments, each segment has the same probability of being selected, and randomly select samples from each segment; calculate the criticality weight of each subsystem and component of the equipment through Latin sampling, and rank the criticality weight of each subsystem and component; On the basis of determining the criticality weight of the equipment, a linear weighted mathematical model is used to calculate the criticality evaluation index of each subsystem and component of the equipment. The calculation formula is: （10） In the formula, Index2 is the hazard evaluation index, f j is the weight of the jth hazard assessment factor; g j is the score of the jth factor, and the subscript j is the serial number of the influencing factor.

Images